1/21
The oxidative stress response: a proteomic view
Thierry Rabilloud1 , Mireille Chevallet1, Sylvie Luche1 and Emmanuelle Leize-Wagner2,+.
1
CEA- Laboratoire d’Immunochimie, DRDC/ICH, INSERM U 548
CEA-Grenoble, 17 rue des martyrs, F-38054 GRENOBLE CEDEX 9, France
2
Laboratoire de Spectrométrie de Masse Bio-Organique, UMR CNRS 7509, ECPM, 25 rue Becquerel,
67008 STRASBOURG Cedex, France
+ present address : Institut de Sciences et ingéniérie Supramoléculaire, UMR CNRS 7006, Université
Louis Pasteur, 8 rue Gaspard Monge, 67083 STRASBOURG CEDEX, France
Correspondence :
Thierry Rabilloud, DRDC/ICH, INSERM U 548
CEA-Grenoble, 17 rue des martyrs,
F-38054 GRENOBLE CEDEX 9
Tel (33)-4-38-78-32-12
Fax (33)-4-38-78-98-03
e-mail: Thierry.Rabilloud@ cea.fr
Abbreviations:
LC : liquid chromatography ; MALDI : Matrix-Assisted Laser Desorption and Ionization ; MS : mass
spectrometry ; MS/MS : tandem mass spectrometry
Running title : proteomics of oxidative stress
Keywords :
Oxidative stress, protein expression, protein carbonyls, proteomics, mass spectrometry, Twodimensional electrophoresis, protein nitration, peroxiredoxin, superoxide dismutase
12/02/2016, 10:02:24
2/21
Abstract
The oxidative stress response is characterized by various effects on a range of biological molecules.
When examined at the protein level, both expression levels and protein modifications are altered by
oxidative stress. While these effects have been studied in the past by classical biochemical methods,
the recent onset of proteomics methods have allowed to study the oxidative stress response on a much
wider scale. The input of proteomics in the study of oxidative stress response or in the evidence of an
oxidative stress component in biological phenomena is thus reviewed in this paper.
12/02/2016, 10:02:24
3/21
1. Introduction
Oxidative stress can be defined as the toxic effect of chemically reactive species derived from oxygen
(superoxide, peroxide, hydroxyl radical) or mixed nitrogen-oxygen species (NO, peroxynitrite). Among
those, the most active species is hydroxyl radical (OH.), most generally produced from hydroperoxides
by the Fenton reaction, using reducing agents and metallic ions able of oscillating between different
valencies (e.g. Cr, Mn, Fe, Ce). This toxic effect can take place on various biological molecules, such
as carbohydrates, unsaturated lipids, proteins, or nucleic acids. Various damages can result on these
substrates. Unsaturated lipids are transformed into lipid peroxides, a reaction which is autoamplified
with molecular oxygen. Proteins undergo various damages. The peptide backbone can be cut by
hydroxyl radical, and many side chains are prone to various modifications. Examples include tyrosine
hydroxylation (by hydroxyl radical), or nitration (by peroxynitrite), methionine or cysteine oxidation
(by hydroxyl radical or peroxide). However, one of the most common modification on protein side
chains is carbonyl formation (while normal proteins do not have any keto or aldehyde groups).
Depending on the nature of the amino acid side chain, this modification can arise by various
mechanisms. While aliphatic side chains (e.g. Val, Leu, Ile) require the very reactive hydroxyl radical
to be oxidized, more reactive side chains (e.g. Trp) can be oxidized by less active oxidants (e.g.
peroxide). There is also some evidence of lysine and arginine oxidation by metal catalyzed oxidation
[1, 2].
Nucleic acids are also targets for oxidative stress-induced damage. Because of the stable nature of
nucleic acids, only hydroxyl radical can oxidize them. One of the most common reaction is the
cleavage of the phosphosugar backbone, in a reaction quite similar to the one used for chemical DNA
footprinting. The other most common reaction is guanine oxidation into 8-oxo guanine, which leads to
mutations at the next replication cycle, as 8-oxoG pairs with adenine rather than cytosine.
It must be recalled, however, that this oxidative stress is common to all organisms living under aerobic
conditions. As all chemical reactions, oxygen reduction under biological conditions is far from being
perfect, and partial reduction with less than four electrons gives rise to reactive oxygen species, which
are direct oxidative stress agents, or can combine with other molecules to yield other oxidative stress
inducers (e.g. superoxide + NO gives peroxynitrite). Consequently, many organisms have evolved
protection or repair mechanism as an adaptation to this constitutive oxidative stress. Examples include
redox enzymes destroying the reactive oxygen species (e.g. superoxide dismutases, catalases,
peroxidases, peroxiredoxins) or chemical reducers encountered in biological systems (e.g. ascorbic
acid, tocopherols), or repair systems removing oxidatively-damaged components of complex molecules
12/02/2016, 10:02:24
4/21
(e.g. lipases, nucleic acid excision systems). The redox enzymes make use of the reducing molecules
produced by the general metabolism (NADH, NADPH) to provide the reducing equivalents needed to
destroy the reactive oxygen species.
Besides this constitutive oxidative stress, cells can be submitted to extra bursts of oxidative stress,
arising from the activity of some enzymes (e.g. oxidases), but also of some signalling pathways (e.g.
TNF).
Thus, because the cells have evolved these protective systems, a real stress occurs only in situations
where the oxidative power present in chemically reactive species overcomes the antioxidant defence
line present in the cells.
Most of our knowledge concerning oxidative stress and the associated cellular responses has been
obtained using classical biochemical approaches. Examples include the role of antioxidant enzymes as
lifespan-prolonging agents in drosophila [3], or the existence of oxidative damages on proteins in
oxidative stress conditions [4], using detection techniques specially designed for the detection of
carbonyl groups on proteins [5]. More generally, the techniques designed to evidence the oxidative
stress-induced modifications on proteins often use immunological detection of modified residues
themselves (e.g. nitrotyrosine) or of haptens that can be selectively coupled to the modified residues
(e.g. detection of carbonyls via immunodetection of coupled dinitrophenylhysrazine). The great
sensitivity of immunodetection allows to detect minute amounts of modified proteins. This is often
required, as the proteasome efficiently degrades oxidatively-damaged proteins, so that proteasome
inhibitors are often used to increase the detection of oxidatively-modified proteins [6].
2. Proteomics studies on oxidative stress response
2.1.General considerations
While these dedicated biochemical techniques have proved useful, they require to test proteins one by
one on an hypothesis-driven basis. Moreover, they focus only on one type of modification and do not
render the complexity of the cellular response to oxidative stress. As a matter of facts, the cellular
response to oxidative stress is not a passive one resulting in the bare accumulation of modified
proteins. Cells submitted to oxidative stress defend themselves by various mechanisms, including
metabolic switches, neosynthesis of proteins to replace the damaged ones, and active degradation of
damaged proteins. Here again, classical approaches based on narrow hypotheses do not reflect the
12/02/2016, 10:02:24
5/21
complexity of the cell response, which is a mix of changes in protein levels, controlled posttranslational modifications of proteins (e.g. phosphorylation) and oxidative damages on proteins.
The onset of global approaches has considerably changed our way of investigating various complex
cellular phenomena. Because oxidative stress response involves many post-translational events,
proteomic approaches are much more suited to the study of this phenomenon than other global
approaches such as transcriptomics.
However, not all proteomics approaches are able to deal correctly with the specific features of the
oxidative stress response. Because oxidative stress is a constitutive phenomenon, mostly quantitative
differences are expected. The proteomics approach chosen must therefore be able to deal with such
quantitative changes. Furthermore, post-translational changes are expected to be a major component of
the oxidative stress response, and the technical approaches chosen must be able to deal with complex
post-translational modifications. It must be emphasized that our knowledge of post-translational
modifications induced by oxidative stress is far from complete. Besides well-known controlled
modifications, such as phosphorylation, the oxidative damages on proteins are not fully described, and
new oxidative modifications are likely to be discovered. Furthermore, we do not have biochemical
tools able to address all known oxidative modifications. For example, there are no antibodies against
oxidation products of tyrosine, tryptophan or methionine.
These constraints rule out certain flavours of proteomics, such as shotgun approaches [7] or isotopecoded, peptide-based approaches [8]. As a matter of facts, most of the proteomics studies of oxidative
stress response published to date have used two-dimensional electrophoresis as a protein separation and
quantification tool, coupled with mass spectrometry as a protein characterization tool.
Despite its weaknesses in the separation of important classes of proteins such as membrane proteins
[9],two-dimensional electrophoresis is still the most performing separation tool when dealing with
complete proteins. Moreover, a rather precise quantitative analysis of relative protein amounts is
possible by this technique. This technique is also able to separate proteins by rather subtle variations of
the isoelectric point (pI) , and many post-translational modifications can be detected by this way. The
most typical example is phosphorylation, but some stages of cysteine oxidation (cysteine sulfinic and
sulfonic acids) are also expected to induce pI changes. Last but certainly not least, the separation power
of two-dimensional gels considerably simplifies the subsequent analysis by mass spectrometry. As a
matter of facts, a spot on a two-dimensional gels contains at most 2-3 proteins. The digestion of these
will produce a few tens of peptides, a complexity that is easily dealt with by peptide mass
fingerprinting approaches based on MALDI-MS or MALDI-MS/MS, or by nanoelectrospray LC/MS
12/02/2016, 10:02:24
6/21
or LC/MS/MS approaches. This low complexity affords in turn an important sequence coverage, which
means that peptides covering an important part of the complete sequence of the proteins of interest are
analyzed in the mass spectrometer. This implies in turn that the likelihood of finding one of the few
modified peptides in a protein altered by oxidative stress is higher with this technical setup than with
other ones dealing with more complex peptide mixtures (e.g. shotgun approaches). Furthermore, a
previous knowledge of the modification is not needed, and MS/MS approaches often allow both the
description of the modification and the localization of the modification in the modified peptide. This
statement does not mean however, that such an assignment is trivial. As a rule of thumb, most
modifications considerably decrease the MS signal. Several mechanisms can take place to explain this
decrease. First of all, many modifications alter the charge of the peptide and often bring negative
charges (e.g. phosphorylation, sulfation, cysteine oxidation in sulfinic or sulfonic acid). This decrease
in turn the ionization efficiency of peptides in cations, which is the usual and optimal ionization mode.
In some extreme cases, it can be even better to take advantage of the strong negative charge and to
analyze peptides in the negative ion mode [10, 11]. Other modifications (e.g. glycosylation) are so
heterogeneous that they split the signal of one peptide into many small peaks. In other cases (e.g.
carbonylation) the modification favors peptide-peptide interactions (e.g. by Schiff base formation),
which decreases in turn the peptide extraction yields and thus the signal. For these reasons,
modification assignment is not easy and poorly documented except for phosphorylation, for which
dedicated strategies have been described (reviewed in [12, 13])
2.2. Studies on prokaryots and lower eukaryots
Examples of proteomics studies of the oxidative stress response cover various biological conditions
and various organisms. Starting with the simplest organisms, i.e. bacteria, such studies cover both
aspects of the oxidative stress response. One aspect is how bacteria resist to a major oxidative stress,
and a good example of this type of study is provided by the analysis of proteins induced by the
presence of hydrogen peroxide in the bacterium Francisella tularensis [14]. This is not only an in vitro
model, as this bacterium survives in macrophages and is thus able to resist to the oxidant species
produced by the NADPH oxidase in the macrophage phagolysosome. The proteins induced in this
bacterium by oxidative stress are typical of a complex oxidative stress response. Besides antioxidant
proteins (AhpC), many chaperones are also induced (to refold proteins unfolded by oxidative stress) ,
as well as proteases (to degrade irreversibly damaged proteins and thus avoid the deleterious effects of
their misfunction). Interestingly enough, the iron metabolism is also changed, in this case by the
12/02/2016, 10:02:24
7/21
induction of a bacterioferritin. More detailed results were also obtained from a more characterized
bacterium, Bacillus subtilis, used as a model organism [15]. Here again, there is a clear induction of
antioxidant enzymes (peroxiredoxins, catalase and superoxide dismutase) and also a clear trend in iron
metabolism aiming at reducing the free iron content in the cell. This can be explained by the fact that
iron react with peroxide to produce hydroxyl radical, which is the most active and the most toxic
oxidant species. The modulation of iron metabolism is thus an important regulation to decrease the
overall toxicity of the peroxide.
Apart from these protein inductions, proteomics has also been used to search for proteins that are
modified by an oxidative stress. Typical examples of such proteins are peroxiredoxins and GAPDH
[16], in which the active site cysteines are oxidized into cysteic acid upon oxidative stress. As this
modification induces a pI shift in the protein, it is easy to find through a 2D electrophoresis-based
approach. Other, more subtle oxidative modifications occurring on cysteines, e.g. disulfide bond
formation cannot be detected as easily and require dedicated approaches [17], using sequential
labelling of thiols prior to or following chemical reduction. These approaches clearly show that several
bacterial proteins undergo disulfide bond formation upon oxidative stress in E coli [17].
Proteomics can also be used in a different way, i.e. to demonstrate an oxidative stress component in a
complex biological response, generally by the observation of an induction of antioxidant proteins. A
good example is provided by the study of the stationary phase in Bacillus licheniformis [18]. This work
clearly shows that many proteins induced by an experimental oxidative stress are also found induced
during the stationary phase. This includes not only antioxidant proteins (catalase and peroxiredoxins)
but also chaperones.
This concept of unicellular model organisms has been extended to eukaryots to further investigate the
oxidative stress response. A very precise work has been made on the yeast Saccharomyces cerevisiae to
identify changes in protein expression induced by peroxides [19]. Interestingly enough, antioxidant
enzymes are not so prominent in this response, which rather emphasises a deep reorientation in the
central metabolism. As an example, glycolysis is decreased while the pentose phosphate pathway is
increased. While glycolysis produces NADH, the pentose phosphate pathway produces NADPH, which
is the reducing substrate used by key antioxidant systems in eukaryots, the glutathione peroxidase
system and the peroxiredoxin system.
2.3.Studies on mammalian systems
The perspective is generally fairly different when the oxidative stress response of mammalian cells is
studied. As a matter of facts, the metabolism of these cells is much less flexible than the one of
12/02/2016, 10:02:24
8/21
bacteria, yeasts or plant cells, so that metabolic adaptation is much less obvious. However, such an
adaptation in the glucide metabolism has been observed in cell lines adapted to survive in low
concentration of hydrogen peroxide [20]. However, The increased abundance of aldose reductase can
also be interpretated as a detoxifying mechanism toward unsaturated or hydroxylated aldehydes, which
are reduced by aldose reductase in the corresponding alcohols. These aldehydes are toxic end products
of the degradation of lipid peroxides, while the corresponding alcohols are much less toxic. There are
only a few studies of this type, examining how mammalian cells can adapt to a constitutive, increased
oxidative stress. A good example is provided for myeloid [21] and epithelial cells [22] . These papers
show that the metabolic adaptation of mammalian cells is much more difficult to understand than the
one of bacteria or yeasts. Besides the induction of some redox regulators (peroxiredoxins in [21] and
[22], superoxide dismutase in [20]) a change in only a few enzymes involved in the glycolysis pathway
was observed. In addition, only a few proteins involved in protein folding are induced.
Besides these studies on the adaptation to oxidative stress, more studies focus on proteins as targets
under oxidative stress conditions. One of the consequences of an oxidative stress may be the decrease
in some protein content in the cell, as a consequence of the oxidative stress damage and of the induced
protein degradation [23]. Here again, it is difficult to draw a clear trend from the variety of proteins that
appear to be decreased upon oxidative stress.
However, most of the proteomics studies on the oxidative stress response of mammalian cells deal with
the wide scale identification of target proteins for a defined modification. There are only a few that try
to identify modifications without a prior hypothesis. Among those, the studies devoted to
peroxiredoxins are a good example of the power of the two-dimensional electrophoresis/ mass
spectrometry approach. The starting point of the study is the simple observation of an altered migration
of some proteins after oxidative stress of the cell. MS analysis first identifies these proteins as the
various members of the peroxiredoxin family, and then shows that the modification is indeed the
overoxidation of the active site cysteine of these proteins [24]. However, this study also shows that
such an approach is not as straightforward as it may seem. The oxidation of cysteine into cysteine
sulfinic or cysteic acid brings a strong negative charge on the modified peptide, thereby greatly
reducing the ionization of the modified peptide in the standard cationic mode used for peptide mass
spectrometry. Consequently, the detailed study of this phenomenon, which has shown that this
modification is reversible in cells [11, 25], has required the development of dedicated mass
spectrometry techniques [10]. Interestingly enough, this cysteine modification is not limited to
peroxiredoxins, and has also been observed in other proteins [26]. This latter work has also
demonstrated other migration alterations upon oxidative stress, which remain to be characterized, as
12/02/2016, 10:02:24
9/21
well as quantitative changes. Moreover, the evidence of the reversibility of this modification [11], [25],
has prompted active research to identify the enzymatic systems responsible for this regeneration [27,
28].
As the thiol group of cysteine is one of the most sensitive redox groups in proteins, other proteomics
studies have aimed at the identification of other thiol modifications upon oxidative stress. The
formation of disulfide bridges has been investigated using off-diagonal electrophoresis [29], and this
resulted in the evidence of complex interprotein bridges. Other studies have aimed at the identification
of disulfide bridges brought upon oxidative stress between protein thiols and glutathione [30]. Last but
not least, a simple but elegant identification methods of free thiol groups in mitochondria has been
described [31]. This method has been recently applied to the identification of oxidized proteins induced
upon alcohol consumption [32], a topic which has also been addressed by other methods involving
differential thiol labelling [33].
However, in all these cases the variety of proteins involved does not allow an easy interpretation of the
results.
Two other common, oxidative stress-induced protein modifications are also frequently assessed in
proteomics studies, namely tyrosine nitration and carbonyl formation. This is due to the availability of
antibodies against the nitrophenyl and dinitrophenyl haptens, which enable a sensitive and specific
detection of these modifications.
These modifications are looked for in two type of studies. In the first type, the cell or tissue of interest
is submitted to a well-characterized oxidative stress, and the target proteins for this stress are
investigated using the modification specific assay by blotting and the resolving power of proteomics.
Examples of such studies have been carried out on muscle [34] or on cells treated with a well-known
oxidative stress inducing toxin [35], and have led to the identification of a few target proteins.
However, precise assignment of the modification sites has not been carried out in these studies.
3. Cell biology proteomics studies related to oxidative stress
Another use of the molecular tools detecting oxidative modifications is to assay these modifications in
biological situations where an oxidative component is suspected. By this way, an increase in the
amount of modified proteins is a hallmark of an oxidative stress. Furthermore, the resolving power of
proteomics allows to identify the target proteins of this oxidative proteins. Such approaches have been
used for tyrosine nitration, e.g. in aging [36], for protein carbonyl formation, e.g. in Alzheimer disease
[37, 38], or in other neurodegenerative disorders [39] or in aging [40, 41] or in apoptosis [42] and for
12/02/2016, 10:02:24
10/21
other oxidative stress-related modifications, such as conjugation with 4 hydroxynonenal [43], which is
a by-product of the oxidative degradation of unsaturated fatty acids
While the expression data are usually difficult to interpret, the data obtained by the study of protein
carbonyls are much simpler, and show an obvious trend for the damage of various chaperone proteins
[37], [42], including mitochondrial chaperones. A review of mitochondrial changes induced by
oxidative stress and detected by proteomics approaches has been recently published [44]. Most often,
the proteomics studies aiming at the identification of oxidatively modified proteins also take into
account the increase or decrease in protein amounts (e.g. [40,42]). However, some studies focus only
on protein amount changes, regardless of any modification-specific study, in biological situations
where oxidative stress is suspected. In this case, it is the changes in antioxidant enzymes which
becomes the hallmark of oxidative stress. Very different biological situations have been investigated by
this approach, including arsenite toxicity [45], the effect of diesel exhaust particles [46],
neurodegeneration [47], brain ischemia [48], chemically-induced cirrhosis [49], aging in liver [50] or
growth factor signalling [51]. The changes are generally detected for peroxiredoxins and glutathione
peroxidases, but sometimes for other proteins such as heme oxygenase [45, 46].
4. Expert opinion
The most important weakness in the proteomics studies of the oxidative stress response is clearly
linked with the limited performance of the current proteomics techniques. Apart from the fact that
membrane proteins are strongly underrepresented [9], which is clearly a problem, especially when
trying to find out target proteins for oxidative stress [44], the representation of the real proteome may
be very variable from one system to another. As a matter of facts, the resolving power of twodimensional gels is 1000-2000 protein spots. When taking into account the chemical diversity brought
by post-translational modifications, this resolving power is close to 50% of the total proteome, and can
reach up to 70% when multiple gels are used [52]. However, it drop down to ca. 10% in mammalian
cells, where many more genes are expressed and where the level of post-translational modifications is
much higher than in prokaryots. This means in turn that our proteomics view of mammalian cells is
very limited to the more soluble and more abundant proteins. This may explain why proteomics results
obtained in oxidative stress response are often so difficult to interpret, as they only provide a very
partial view of the whole picture.
Alternate methods such as shotgun proteomics, based on multidimensional chromatographic separation
of peptides [7], does not present the same drawbacks and is able to analyze many more peptides with a
12/02/2016, 10:02:24
11/21
much less bias against membrane proteins. However, this approach is very difficult to make
quantitative and offers limited sequence coverage when mixtures as complex as whole cell extracts are
analyzed. This makes these methods less suited for analysis of oxidative stress, which is mostly a
quantitative phenomenon in which post-translational modifications play important role, which makes
the sequence coverage a crucial parameter.
Apart from this classical representation problem in proteomics, which are a bottleneck in all
proteomics studies, another more specific problem arises in the study of the oxidative stress. Many
interesting modifications on target proteins arise from oxidative damages, which can take very different
forms. One of the most documented modifications is carbonyl formation. This modification is usually
detected by conjugation of the carbonyl group with an hydrazine or hydrazide, and the resulting
hydrazone is assayed by detection of the chemical moiety coupled with the hydrazine group. This can
be by an antibody-hapten reaction, or by avidin-biotin affinity [53]. However, this process is far from
being perfect, as the hydrazone formation is very pH-dependent and easily decreased by steric or
electrostatic interactions. Furthermore, the excess reactant must be removed by precipitation of
proteins. This process induces protein losses and also partial cleavage of the hydrazone.
While other modifications such as tyrosine nitration are efficiently detected by the way of specific
antibodies, there are however many other modifications for which no molecular tools are available.
This is for example the case of tyrosine hydroxylation, but also of methionine or tryptophan oxidation.
It is also very likely that some modifications remain to be discovered.
In this scenario, the pI shift that is often associated with oxidative modifications may be a good tool to
separate the modified forms from the bulk of unmodified proteins. The resolving power of narrowrange pH range should improve the number of cases in which a modified protein will be separated from
the unmodified one. It is then crucial to achieve an optimal sequence coverage in mass spectrometry to
maximize the probability of identifying the modification and assigning it as precisely as possible
ideally to a single amino-acid (e.g. in [24]). This is clearly an important area of improvement over most
of the studies published to date, where the modifications are most often just assigned to a protein and
not more precisely than that.
5. Five-year view
The expectations that ca be put five years ahead lie on the directions outlined in the previous
paragraph. On the one hand, the continuous progress in protein separations will increase our
representation of proteomes in general, and this will of course have an impact in the studies of
12/02/2016, 10:02:24
12/21
oxidative stress. This also holds true for the mass spectrometry side, which becomes always more
precise and comprehensive. A quantum leap in the precision of the description of the oxidative stressinduced modifications is very likely in the next few years, aiming at assigning in most cases the precise
site of modification. Together with improved structural knowledge or predictions in protein structure,
this will considerably help in understanding how modifications modulate protein function. Her again,
peroxiredoxins provide a very good example, as the normal on oxidized form show neither the same
structure nor the same function [54].
The situation of the molecular tools, and especially antibodies, for detecting modifications is however
much less encouraging, and it is very likely that the modifications escaping detection today will still
not be detected by this means within the next few years.
However, one recently-introduced technique should be very useful for the study of oxidative stressinduced modifications. This technique has been nicknamed COFRADIC, for Combined FRACtions
DIagonal Chromatography [55]. The rationale of the technique is to select peptides in a complex
mixture on the basis of a chemical property. To this purpose, the complex peptide mixture is first
fractionated by reverse phase chromatography, and fractions are collected. All fractions are then
submitted to a chemical treatment that will specifically modify some peptides in the mixture. Each
fraction is then refractionated on the same column than in the first
fractionation. The bulk of
unmodified peptide will elute at the same position in the gradient and these are discarded. The
modified peptides will elute either earlier or later in the gradient, depending if the chemical treatment
makes them more hydrophilic or more hydrophobic. Those peptides are then selectively analyzed in the
mass spectrometer. This approach has been recently reviewed and described in detail [56]. The great
interest of this method is its versatility, and many different modifications schemes can be devised to
detect classes of peptides. Interestingly enough, the proof of concept of the method has been made on
methionine-containing peptides, and the chosen modification was methionine oxidation with hydrogen
peroxide [55]. This means conversely that peptides containing oxidized methionine can be selectively
analyzed by making a reduction between the two separations, e.g. with borohydride. Carbonylcontaining peptides should also be selectable, e.g. with reductive amination with cyanoborohydride, or
hydrazone formation with a less bulky hydrazine than the usual dinitrophenylhydrazine.
The versatility of organic chemistry should greatly help in designing modification schemes able to
select many of the oxidative modifications that are today difficult to analyze (e.g. tyrosine or
tryptophan oxidation, but also oxidation at other amino acids). Furthermore, the concept can be
increased to an even greater flexibility by investigating the usefulness of peptide cleavages. Up to now,
rather subtle modifications have been used. They do not modify too much the overall properties of the
12/02/2016, 10:02:24
13/21
peptides, so that some fractions can be pooled in the second analysis (hence the name COmbined
FRActions). For example, fractions 1, 11,21 etc, can be analyzed in a single run, as it is very unlikely
that a methionine oxidation or a phosphate removal will move one peptide as wide as ten fractions.
However, if the experimentators are prepared to run fractions one by one, then a peptide cleavage,
which will induce dramatic changes in the chromatographic mobility between the parent peptides and
the cleavage peptides, can be used. As to oxidative stress, this could be very useful to analyze oxidation
of aromatic amino acids. As a matter of fact, many :classical methods for chemical cleavage after
aromatic amino acids (Tyr or Trp), start with the oxidation of the amino acids, followed by a cleavage
step that uses the reactivity of the oxidized amino acid [57, 58]. The use of the sole cleavage step could
provide an efficient tool to localize oxidized aromatic amino acids.
As for all peptide-based proteomics approaches, it is difficult to perform a quantitative analysis with
this method. However, it is possible to devise a two-stage approach allowing a relative quantification
of the modified peptide. The first stage consists in the determination of the modified peptide with the
COFRADIC approach. Once the peptide has been identified and characterized, it is possible to predict
via sequence analysis the mass of the unmodified peptide. In the second stage, a standard LC/MS
analysis then allows to reach the intensity ratio between the MS signals for the modified vs.
unmodified peptide. If the modification is discrete enough, it should minimally alter the ionization
efficiency for the modified peptide compared to the normal one, so that the signal ratio between the
two is a good approximation of the real quantitative ratio.
In conclusion, the coming years should demonstrate a major improvement in the description of the
oxidative stress-induced modifications of proteins, both in the variety of modifications assessed and in
the precision of the determination of the modified sites. This should provide much more precise
insights in the deleterious effects of oxidative stress on proteins. As a matter of facts, it should be kept
in mind that the data obtained by proteomics on protein modifications are just a first step. The
following one is to demonstrate that the modifications observed on a protein alter its activity in a way
or another, either directly by altering its conformation and/or the active site (e.g. cysteine oxidation in
the active site of peroxiredoxins or GAPDH), or by altering the association to other proteins. Such
rigorous demonstrations should also considerably enhance our comprehension of the deleterious
phenomena occurring during oxidative stress
6. Key issues
-Proteomic studies allow to study the oxidative stress response with a much wider view than
12/02/2016, 10:02:24
14/21
conventional biochemical methods.
-Proteomic studies also allow to evidence the oxidative stress component in various complex cellular
processes, including cell signalling, aging and pathologies
-With the development of dedicated biochemical tools and their adaptation in the proteomics toolbox, it
is now possible to study both protein expression variations and oxidative protein modifications (e.g.
tyrosine nitration or carbonyl formation) by the wide-screen proteomics toolbox
-The current proteomics toolbox is not comprehensive enough, however, to deliver a clear picture of
the complex cellular events taking place upon oxidative stress. The data obtained to date are generally
very fragmented.
-Despite this partial character, the importance of subclasses of antioxidant enzymes (e.g.
peroxiredoxins) and of chaperone proteins, has been exemplified in numerous studies of the oxidative
stress response.
-Besides the protein analysis scope, which is a major challenge in proteomics and which will be
beneficial to all proteomics studies, the study of the oxidative stress response would benefit of
dedicated improvements, especially in the detection of nowadays poorly characterized oxidative
modifications (e.g. tyrosine hydroxylation)
12/02/2016, 10:02:24
15/21
References
*[1] Requena JR, Chao CC, Levine RL, Stadtman ER. Glutamic and aminoadipic
semialdehydes are the main carbonyl products of metal-catalyzed oxidation of proteins.
Proc Natl Acad Sci U S A. 98(1):69-74. (2001).
Indispensable to understand oxidative protein carbonyl formation
[2] Requena JR, Stadtman ER.
Conversion of lysine to N(epsilon)-(carboxymethyl)lysine increases
susceptibility of proteins to metal-catalyzed oxidation.
Biochem Biophys Res Commun. 264(1):207-211 (1999).
[3] Orr WC, Mockett RJ, Benes JJ, Sohal RS. Effects
of overexpression of copperzinc and manganese superoxide dismutases, catalase, and thioredoxin
reductase genes on longevity in Drosophila melanogaster. J Biol Chem.
278(29):26418-26422 (2003).
[4]Yan LJ, Levine RL, Sohal RS. Oxidative damage during aging targets mitochondrial
aconitase.Proc Natl Acad Sci U S A. 94(21):11168-11172 (1997).
[5] Levine RL, Williams JA, Stadtman ER, Shacter E. Carbonyl assays for determination of
oxidatively modified proteins. Methods Enzymol. 233:346-357 (1994)
[6] Drake SK, Bourdon E, Wehr NB, Levine RL, Backlund PS, Yergey AL, Rouault TA.
Numerous proteins in Mammalian cells are prone to iron-dependent oxidation and
proteasomal degradation.Dev Neurosci. 24(2-3):114-124 (2002).
[7] Washburn MP, Wolters D, Yates JR 3rd. Large-scale
analysis of the yeast
proteome by multidimensional protein identification technology. Nat
Biotechnol. 19(3):242-247 (2001).
[8] Gygi SP, Rist B, Gerber SA, Turecek F, Gelb MH, Aebersold R. Quantitative
analysis
of complex protein mixtures using isotope-coded affinity tags. Nat Biotechnol.
17(10):994-999 (1999).
[9] Santoni V, Molloy M, Rabilloud T. Membrane
proteins and proteomics: un
amour impossible? Electrophoresis. ;21(6):1054-1070 (2000)
[10] Wagner E, Luche S, Penna L, Chevallet M, Van Dorsselaer A, Leize-Wagner E, Rabilloud
T. A
method for detection of overoxidation of cysteines: peroxiredoxins are
oxidized in vivo at the active-site cysteine during oxidative stress. Biochem J.
366(Pt 3):777-785 (2002).
12/02/2016, 10:02:24
16/21
[11] Chevallet M, Wagner E, Luche S, van Dorsselaer A, Leize-Wagner E, Rabilloud T.
Regeneration of peroxiredoxins during recovery after oxidative stress: only
some overoxidized peroxiredoxins can be reduced during recovery after
oxidative stress. J Biol Chem. 278(39):37146-37153 (2003).
[12] Reinders J, Sickmann A.
State-of-the-art in phosphoproteomics.
Proteomics. (2005) in press
[13] Areces LB, Matafora V, Bachi A.
Analysis of protein phosphorylation by mass spectrometry.
Eur J Mass Spectrom;10:383-392 (2004).
[14] Lenco J, Pavkova I, Hubalek M, Stulik J.Insights into the oxidative stress response in
Francisella tularensis LVS and its mutant DeltaiglC1+2 by proteomics analysis.
FEMS Microbiol Lett. 246(1):47-54 (2005)
*[15] Mostertz J, Scharf C, Hecker M, Homuth G. Transcriptome and proteome analysis of
Bacillus subtilis gene expression in response to superoxide and peroxide stress.
Microbiology. 150(Pt 2):497-512 (2004).
A typical study on a prokaryot, examplifying the analysis detail
[16] Weber H, Engelmann S, Becher D, Hecker M.
Oxidative stress triggers thiol oxidation in the glyceraldehyde-3-phosphate
dehydrogenase of Staphylococcus aureus.
Mol Microbiol. 52:133-140. (2004)
[17] Leichert LI, Jakob U.
Protein thiol modifications visualized in vivo.
PLoS Biol. 2:e333 (2004).
[18] Voigt B, Schweder T, Becher D, Ehrenreich A, Gottschalk G, Feesche J, Maurer KH,
Hecker M.A proteomic view of cell physiology of Bacillus licheniformis. Proteomics. 4(5):14651490 (2004).
*[19] Godon C, Lagniel G, Lee J, Buhler JM, Kieffer S, Perrot M, Boucherie H, Toledano MB,
Labarre J. The H2O2 stimulon in Saccharomyces cerevisiae.
J Biol Chem. 273(35):22480-22489 (1998).
A pioneer oxidative stress sudy by proteomics, showing the complex metabolism reorientation
[20] Keightley JA, Shang L, Kinter M. Proteomic analysis of oxidative stress-resistant cells: a
specific role for aldose reductase overexpression in cytoprotection. Mol Cell Proteomics.
12/02/2016, 10:02:24
17/21
3(2):167-175 (2004).
[21] Seong JK, Kim do K, Choi KH, Oh SH, Kim KS, Lee SS, Um HD. Proteomic analysis of
the cellular proteins induced by adaptive concentrations of hydrogen peroxide in human U937
cells. Exp Mol Med. 34(5):374-378 (2002).
[22] Chan HL, Gharbi S, Gaffney PR, Cramer R, Waterfield MD, Timms JF.
Proteomic analysis of redox- and ErbB2-dependent changes in mammary luminal epithelial
cells using cysteine- and lysine-labelling two-dimensional difference gel electrophoresis.
Proteomics. 5 : 2908-2926 (2005)
[23] Vorum H, Ostergaard M, Hensechke P, Enghild JJ, Riazati M, Rice GE. Proteomic
analysis of hyperoxia-induced responses in the human choriocarcinoma cell line JEG-3.
Proteomics. 4(3):861-867 (2004).
*[24] Rabilloud T, Heller M, Gasnier F, Luche S, Rey C, Aebersold R, Benahmed M, Louisot
P, Lunardi J. Proteomics analysis of cellular response to oxidative stress.
Evidence for in vivo overoxidation of peroxiredoxins at their active site.
J Biol Chem. 277:19396-19401(2002).
How to detect a new oxidative modification with the proteomics toolbox
[25] Woo HA, Chae HZ, Hwang SC, Yang KS, Kang SW, Kim K, Rhee SG.
Reversing the inactivation of peroxiredoxins caused by cysteine sulfinic acid
formation.Science. 300(5619):653-656 (2003).
[26] Paron I, D'Elia A, D'Ambrosio C, Scaloni A, D'Aurizio F, Prescott A, Damante G, Tell G. A
proteomic approach to identify early molecular targets of oxidative stress in human epithelial
lens cells. Biochem J. 378(Pt 3):929-937 (2004)
[27] Biteau B, Labarre J, Toledano MB. ATP-dependent
reduction of cysteine-
sulphinic acid by S. cerevisiae sulphiredoxin. Nature. 425(6961):980-984 (2003).
[28] Budanov AV, Sablina AA, Feinstein E, Koonin EV, Chumakov PM.
Regeneration of peroxiredoxins by p53-regulated sestrins, homologs of
bacterial AhpD.Science. 304(5670):596-600 (2004).
[29] Brennan JP, Wait R, Begum S, Bell JR, Dunn MJ, Eaton P. Detection and mapping of
widespread intermolecular protein disulfide formation during cardiac oxidative stress using
proteomics with diagonal electrophoresis. J Biol Chem. 279(40):41352-41360 (2004)
*[30] Fratelli M, Demol H, Puype M, Casagrande S, Eberini I, Salmona M, Bonetto V,
Mengozzi M, Duffieux F, Miclet E, Bachi A, Vandekerckhove J, Gianazza E, Ghezzi P.
Identification by redox proteomics of glutathionylated proteins in oxidatively stressed human T
lymphocytes. Proc Natl Acad Sci U S A. 99(6):3505-3510 (2002).
How to detect a new oxidative modification by the proteomics toolbox
12/02/2016, 10:02:24
18/21
[31] Lin TK, Hughes G, Muratovska A, Blaikie FH, Brookes PS, Darley-Usmar V, Smith RA,
Murphy MP. Specific modification of mitochondrial protein thiols in response to oxidative
stress: aproteomics approach. J Biol Chem. 277(19):17048-17056 (2002).
[32] Venkatraman A, Landar A, Davis AJ, Ulasova E, Page G, Murphy MP, Darley-Usmar V,
Bailey SM.
Oxidative modification of hepatic mitochondria protein thiols: effect of
chronic alcohol consumption.
Am J Physiol Gastrointest Liver Physiol. 286:G521-G527 (2004)
[33] Suh SK, Hood BL, Kim BJ, Conrads TP, Veenstra TD, Song BJ.
Identification of oxidized mitochondrial proteins in alcohol-exposed human
hepatoma cells and mouse liver.
Proteomics. 4:3401-3412. (2004)
[34] Stagsted J, Bendixen E, Andersen HJ. Identification of specific oxidatively modified proteins in
chicken muscles using a combined immunologic and proteomic approach. J Agric Food Chem.
52(12):3967-3974 (2004).
[35] Pocernich CB, Poon HF, Boyd-Kimball D, Lynn BC, Nath A, Klein JB, Butterfield DA.
Proteomic analysis of oxidatively modified proteins induced by the mitochondrial toxin 3nitropropionic acid in human astrocytes expressing the HIV protein tat.
Brain Res Mol Brain Res. 18;133(2):299-306 (2005).
[36] Kanski J, Hong SJ, Schoneich C.
Proteomic analysis of protein nitration in aging skeletal muscle and identification of
nitrotyrosine-containing sequences in vivo by nanoelectrospray ionization tandem mass
spectrometry. J Biol Chem.280(25):24261-24266 (2005)
*[37] Korolainen MA, Goldsteins G, Alafuzoff I, Koistinaho J, Pirttila T.
Proteomic analysis of protein oxidation in Alzheimer's disease brain.
Electrophoresis. 23(19):3428-3433 (2002).
the oxidative stress component of a pathology
[38]Castegna A, Thongboonkerd V, Klein JB, Lynn B, Markesbery WR, Butterfield DA.
Proteomic identification of nitrated proteins in Alzheimer's disease brain.
J Neurochem. 2003 Jun;85(6):1394-401.
[39] Poon HF, Hensley K, Thongboonkerd V, Merchant ML, Lynn BC, Pierce WM, Klein JB,
Calabrese V, Butterfield DA.
Redox proteomics analysis of oxidatively modified proteins in G93A-SOD1
transgenic mice--a model of familial amyotrophic lateral sclerosis.
Free Radic Biol Med. 39:453-462 (2005)
12/02/2016, 10:02:24
19/21
[40] Poon HF, Vaishnav RA, Getchell TV, Getchell ML, Butterfield DA.
Quantitative proteomics analysis of differential protein expression and oxidative modification
of specific proteins in the brains of old mice. Neurobiol Aging. (2005) in press
[41] Poon HF, Castegna A, Farr SA, Thongboonkerd V, Lynn BC, Banks WA, Morley JE,
Klein JB, Butterfield DA. Quantitative proteomics analysis of specific protein expression and
oxidative modification in aged senescence-accelerated-prone 8 mice brain.
Neuroscience. 126(4):915-926 (2004)
*[42] Magi B, Ettorre A, Liberatori S, Bini L, Andreassi M, Frosali S, Neri P, Pallini V, Di
Stefano A.Selectivity of protein carbonylation in the apoptotic response to oxidative stress
associated with photodynamic therapy: a cell biochemical and proteomic investigation.
Cell Death Differ. 11(8):842-852 (2004).
A typical integrated proteomics study of oxidative stess in mammalian cells
[43] Perluigi M, Fai Poon H, Hensley K, Pierce WM, Klein JB, Calabrese V, De Marco C,
Butterfield DA.
Proteomic analysis of 4-hydroxy-2-nonenal-modified proteins in G93ASOD1 transgenic mice--a model of familial amyotrophic lateral sclerosis.
Free Radic Biol Med. 38:960-968 (2005)
[44] Bailey SM, Landar A, Darley-Usmar V.Mitochondrial proteomics in free radical research.
Free Radic Biol Med. 38(2):175-188. (2005).
[45] Lau AT, He QY, Chiu JF.A proteome analysis of the arsenite response in cultured lung
cells: evidence for in vitrooxidative stress-induced apoptosis. Biochem J. 382(Pt 2):641-650
(2004).
[46] Xiao GG, Wang M, Li N, Loo JA, Nel AE. Use of proteomics to demonstrate a
hierarchical oxidative stress response to diesel exhaust particle chemicals in a macrophage
cell line. J. Biol. Chem. 50781-50790 (2003).
[47] Krapfenbauer K, Engidawork E, Cairns N, Fountoulakis M, Lubec G.
Aberrant expression of peroxiredoxin subtypes in neurodegenerative disorders.
Brain Res. 967(1-2):152-160 (2003).
[48] Schrattenholz A, Wozny W, Klemm M, Schroer K, Stegmann W, Cahill MA.
Differential and quantitative molecular analysis of ischemia complexity reduction by isotopic
labeling of proteins using a neural embryonic stem cell model. J Neurol Sci. 15;229-230:261267 (2005).
[49] Low TY, Leow CK, Salto-Tellez M, Chung MC. A proteomic analysis of thioacetamideinduced hepatotoxicity and cirrhosis in rat livers.
Proteomics. 4(12):3960-3974 (2004).
[50] Cho YM, Bae SH, Choi BK, Cho SY, Song CW, Yoo JK, Paik YK. Differential expression
of the liver proteome in senescence accelerated mice. Proteomics. 3(10):1883-1894 (2003).
12/02/2016, 10:02:24
20/21
[51] Imamura T, Kanai F, Kawakami T, Amarsanaa J, Ijichi H, Hoshida Y, Tanaka Y, Ikenoue
T, Tateishi K, Kawabe T, Arakawa Y, Miyagishi M, Taira K, Yokosuka O, Omata M.
Proteomic analysis of the TGF-beta signaling pathway in pancreatic
carcinoma cells using stable RNA interference to silence Smad4 expression.
Biochem Biophys Res Commun. 318(1):289-296 (2004)
[52] Tonella L, Hoogland C, Binz PA, Appel RD, Hochstrasser DF, Sanchez JC.
New perspectives in the Escherichia coli proteome investigation.
Proteomics. 1(3):409-423 (2001).
[53] Soreghan BA, Yang F, Thomas SN, Hsu J, Yang AJ.
High-throughput proteomic-based identification of oxidatively induced protein carbonylation in
mouse brain.
Pharm Res. 20:1713-1720 (2003)
[54] Jang HH, Lee KO, Chi YH, Jung BG, Park SK, Park JH, Lee JR, Lee SS, Moon JC, Yun
JW, Choi YO, Kim WY, Kang JS, Cheong GW, Yun DJ, Rhee SG, Cho MJ, Lee SY.
Two enzymes in one; two yeast peroxiredoxins display oxidative stressdependent switching from a peroxidase to a molecular chaperone function.
Cell. 117(5):625-635 (2004).
*[55] Gevaert K, Van Damme J, Goethals M, Thomas GR, Hoorelbeke B, Demol H, Martens L,
Puype M, Staes A, Vandekerckhove J.Chromatographic
isolation of methioninecontaining peptides for gel-free proteome analysis: identification of more
than 800 Escherichia coli proteins.Mol Cell Proteomics. 1(11):896-903 (2002).
*the potentiality and versatility of the COFRADIC approach
*[56] Gevaert K, Damme PV, Martens L, Vandekerckhove J.
Diagonal reverse-phase chromatography applications in peptide-centric
proteomics: Ahead of catalogue-omics?
Anal Biochem. 345:18-29 (2005)
*the potentiality and versatility of the COFRADIC approach
[57] Huang HV, Bond MW, Hunkapiller MW, Hood LE.
Cleavage at tryptophanyl residues with dimethyl sulfoxide-hydrochloric acid
and cyanogen bromide.
Methods Enzymol. 91:318-324. (1983)
[58] Savige WE, Fontana A.Cleavage
of the tryptophanyl peptide bond by
dimethyl sulfoxide-hydrobromic acid.
Methods Enzymol. 47:459-469 (1977)
12/02/2016, 10:02:24
21/21
12/02/2016, 10:02:24